机械臂带电作业中一种基于目标识别的手眼标定方法

A Hand-eye Calibration Method Based on Target Recognition in Mechanical Arm Live Work

电网运维检修可使用手眼标定技术通过相机引导带电作业机械臂对输、变电装备进行识别定位和作业,但传统手眼标定方法存在手眼位姿相对固定的问题,因此,文章提出一种基于3D标识球和融合惯性测量单元(Inertial Measurement Unit,IMU)的双目相机手眼关系外参数标定方法。所提方法以3D打印球作为标识物,可快速标定机械臂基坐标系与相机坐标系之间的位置姿态转换矩阵,当相机位姿改变时,通过IMU返回的惯导数据对相机位姿进行修正,从而解决相机位姿变化时需重新标定的问题。仿真实验表明,所提方法误差小于IMU惯导误差。手眼标定实验结果表明,实测标定结果与仿真实验结果一致。

Hand-eye calibration technology can guide the camera to identify and operate the transmission and transformation equipment,while the traditional hand-eye calibration method has a relatively fixed hand-eye posture.Therefore,we propose an external parameter calibration method for a binocular camera based on the 3D identification ball and the fusion inertial measurement unit(Inertial Measurement Unit,IMU).The proposed method uses the 3D printing ball as the identification object,which can quickly calibrate the position and posture conversion matrix between the robot base coordinate system and the camera coordinate system.When the camera position changes,the camera position is corrected through the habitual guide data returned by IMU,thus the problem of the camera position change is solved.In the simulation experiments,the error does not exceed the habitual guide error of the IMU.In the hand-eye calibration experiment,the measured calibration results are consistent with the simulation experiment.